Red Sun Setting: The Battle of the Philippine Sea

Introduction

BY JUNE OF 1944 the tide that had swept the Japanese forces across the Pacific to Wake, Guadalcanal, and New Guinea had begun to ebb rapidly. The Pacific War had begun impressively enough for the Japanese with a series of stunning victories at Pearl Harbor, Hong Kong, Singapore, in the Java Sea and the Indian Ocean, and in the Philippines. By these victories the Japanese had forced the U.S. Navy to rethink its strategic and tactical options. The battleship, beloved vessel of the Big Gun admirals, was pushed to one side and the aircraft carrier became the new capital ship. Unfortunately, United States carrier tactics were still very much in the formative stage. In a battleship-oriented navy, the tactics, administration, and other policies that make a carrier unit run efficiently were still being formulated when 7 December 1941 came. Then there was no time to go over procedures systematically; the carriers were sent out to fight. But by June 1944 the concepts, procedures, and policies had all fallen into place. Of this the Japanese Navy was only too aware.

The awesome victories for the Japanese continued for some months, but then came Coral Sea, Midway, and the furious series of actions on and around Guadalcanal. Besides the precious carriers that were lost in some of these battles, the Japanese had lost irreplaceable flight crews. Their airmen had been the best trained in the Pacific during the early days of the war, but they had been small in numbers. The Japanese had planned on quick conquests, using a small elite group of fliers. They had not planned to augment this group with additional trained airmen at a later time. As a result, few qualified fliers were available to take up the slack caused by the losses in these crucial battles, and it would be some time before a training program would become effective. The need for experienced airmen would become all too apparent to the Japanese military leaders in the months ahead. Indeed, lower-ranking Japanese would not be the only ones to fall in combat. On 18 April 1943 Admiral Isoroku Yamamoto’s plane was ambushed by United States Army Air Force planes and the admiral was killed. As with Midway, this victory could be credited to the marvelous work of the American cryptanalysts.

By the end of 1943 the Solomons campaign was almost over and strategic interest now switched to the Central Pacific. In November 1943 the Gilberts were taken by the Americans. Heavy losses were sustained by the Marines at Tarawa, but by this point in the war the Americans were beginning to digest the lessons they had been force-fed by the Japanese. On 6 January 1944 the Fast Carrier Task Force (initially built around a core of six fleet and five light carriers but constantly increasing in size) became Task Force 58 under the command of Rear Admiral Marc A. Mitscher.

Task Force 58 began a series of attacks—the Marshalls, Truk, the Marianas, Palau, Hollandia—that kept the Japanese guessing and subtracted more aircraft and fliers from their order of battle. In May TF 58 was anchored at Majuro, in the Marshalls, for a well-earned rest. A big operation was just around the corner—Forager, the invasion of the Marianas. Not since the big fleet actions around Guadalcanal in 1942 had the Japanese sought a large-scale battle. Some Americans thought they never would. Others thought that with the right provocation the Japanese Navy would come out spoiling for a fight. It was hoped that Forager be the right provocation.

chapter 1

A Long Winding Road

UNITED STATES STRATEGIC PLANNING for Pacific operations during World War II can be likened to a tortuous back road in the mountains. Along the winding road on the way to Tokyo were a number of stops and an occasional side road. One of these stops, in the Marianas, did supply the provocation for the Japanese Fleet to seek a naval action with the United States Fifth Fleet. However, the road to the Marianas engagement was not a smooth one, and for some time thought was given to taking a side road and bypassing these islands entirely.

Even before the war, the Marianas had figured importantly in U.S. naval plans for the Pacific. In the event of a war with Japan the pre-World War II Orange and Rainbow plans called for U.S. forces to move across the Pacific via the Marshalls and the Marianas to the Philippines.¹ It was assumed the latter islands would be under heavy attack or even lost. American forces would move through the Central Pacific to drive off the Japanese and relieve the Philippines.

The Pearl Harbor attack and the stunning early victories of the Japanese in the Pacific threw out of balance these pre-war plans. The direction of the Japanese attacks also tended to color U.S. strategic planning in the first years of the war. With Japanese and United States forces fighting in the South and Southwest Pacific during this time, the attention of the U.S. strategists was drawn in that direction.² Operations in the Central Pacific were limited to weak raids on widely scattered targets. Still, the traditional Navy view of war against Japan in any case was that the major offensive would be across the Central Pacific rather than through the East Indies.³

Fortunately, the war plans issued were not cast in concrete. Circumstances changed and so plans changed. The operations against the Marianas showed just how flexible these plans could be.

At the time of the Casablanca Conference in January 1943 there was no final, approved plan in existence for the defeat of Japan.⁴ In August of the previous year the Joint United States Strategic Committee had begun work on a strategic plan for the defeat of Japan, but this plan was far from finished when the American and British leaders met at Casablanca. Although nothing was yet in solid form, Admiral Ernest J. King, Commander in Chief, U.S. Fleet and Chief of Naval Operations, undertook to present his views on the Pacific situation.

King’s ideas at this stage of the war closely followed the concept developed in the preceding years of war games at Newport and of successive plans named Orange.⁵ An American advance, according to King, should be toward the Philippines—but by way of the Marshalls, Truk, and the Marianas, not via New Guinea and the Netherlands East Indies. These latter areas were not, in King’s view, the proper places for the use of American naval forces. King stressed the Marianas as the key of the situation because of their location on the Japanese line of communications.⁶

So it was here at Casablanca that King described to the Combined Chiefs of Staff (CCS) the line of advance through the Central Pacific to the Philippines that was in fact to be the primary strategic pattern for the war against Japan.⁷ King was sure that America’s growing naval strength (which would see in 1943 a massive increase with the addition of the new Essex-class carriers and Iowa-class battleships) would eventually force the Japanese into submission without the terrible losses which would be incurred in an invasion of Japan.⁸ It took many months of high-level wrangling before a Central Pacific route was approved, but King’s confidence in his naval forces was well founded. When the fast carriers showed they could operate without land-based air cover, deep in enemy territory, the Central Pacific drive moved more rapidly and deeply into the Japanese defenses.⁹

It would take some time for King’s ideas to be digested by the Combined Chiefs. In the meantime, General Douglas MacArthur, Commander in Chief Southwest Pacific, found King’s plans thoroughly indigestible and wasted no time in criticizing it. In his campaign plan for the Southwest Pacific, Reno I issued in February 1943, he claimed that the route King favored would be time consuming and expensive in . . . naval power and shipping.¹⁰ A drive up the back of New Guinea and into the Philippines (under his command, of course) would be more successful.

MacArthur’s protests had little effect. In March 1943, representatives of the three major areas (South, Southwest, and Central) met in Washington for a Pacific Military Conference. Most of the agenda was taken up with operations in the South and Southwest Pacific and the endless debate over the division of resources between Europe and the Pacific. However, King and Admiral Chester W. Nimitz’s representative, Vice Admiral Raymond A. Spruance, were able to put in a few words for a Central Pacific offensive. King pointed out indirectly that major Navy units could be more valuable in the Central Pacific than in the Solomons—New Guinea area. Spruance then pointed out that with Japanese naval forces still afloat, Pearl Harbor remained a tempting target. Ships from the South Pacific, along with the new vessels just becoming available, could launch an assault on the Gilberts and Marshalls, and remove the threat of an attack on Hawaii.¹¹ (Left unsaid but very likely considered by King was the thought that a Gilberts or Marshalls operation would provide the opening for the Navy to push through to the Joint Chiefs of Staff (JCS) more reasons for a Central Pacific offensive.)

At the Trident Conference held in Washington in May 1943, a Central Pacific route was tentatively approved by the CCS. In the Strategic Plan for the Defeat of Japan presented to and approved by the Combined Chiefs, however, it appeared that the planners were leaning toward MacArthur’s views. In this plan Nimitz’s forces were given a secondary role of capturing the Gilberts, Marshalls, and Truk, while protecting MacArthur’s right flank. MacArthur, in turn, would be rolling up New Guinea into the Celebes and Sulu Sea areas.¹² But MacArthur was still upset that any planning would be considered for a Central Pacific drive. Nevertheless, he did not argue too strongly against the invasions of the Gilberts and the Marshalls since they would remove a threat to his flank. Any further moves through the Central Pacific would be another matter, however.¹³

In the meantime King was still fighting to get the Marianas targeted by the CCS. At the Trident Conference King tried once again. His intense interest in the Marianas stemmed from a realization on his part that the true importance of this target was not unanimously felt. Significant in this connection is the fact that the Marianas had not been prescribed as a specific objective even at this time.¹⁴

For the Navy to undertake an offensive in the Central Pacific, major naval units would have to be transferred from the South Pacific to join with the new construction just appearing. MacArthur and Vice Admiral William F. Halsey, Commander in Chief South Pacific, definitely opposed this idea as did one of Nimitz’s deputies, Vice Admiral John H. Towers. As Commander Air Force Pacific Fleet, Towers believed that Rabaul (still in the plans to be assaulted) was the best Japanese base for American use in operations against enemy bases in the Carolines and sea routes between Japan and Malaysia.¹⁵ Nevertheless, Admiral King was for the redeployment of certain naval units, and this movement did take place.

Following the Trident meetings King reviewed the plans that had been proposed and urged that a timetable be set up for Central Pacific operations. Shortly the Joint Staff Planners agreed to prepare an outline plan for the capture of the Marshalls in November of 1943. Nimitz was directed by the JCS to submit an operational plan and target date for the Marshalls. A tentative plan was received in July.¹⁶

The Joint Strategic Survey Committee presented to the JCS in late June a memorandum covering Pacific operations to that time and making several recommendations for the future conduct of the war. In the Committee’s view the war against Japan had fallen into a strategy of attack from the South and Southwest Pacific because the defense of Australia had required so much of the available resources. This, coupled with certain psychological and political considerations¹⁷ and a lack of U.S. naval forces, had tended to spotlight an offensive strategy from Australia. It had originally seemed that an offensive in the Bismarcks—Solomons—New Guinea area would be fairly successful. This turned out not to be the case, however. Strong Japanese bases in the area had made the offensive slow and relatively costly. But now it appeared that, with U.S. naval strength growing, a drive through the Central Pacific, supported by these naval forces, would be more successful and offer greater strategic advantages.¹⁸

By this time in the war, after the losses in the Solomons and the Aleutians, the Japanese began to worry that they were overextended. And there was a gnawing fear of an American offensive in the Central Pacific. Thus the Japanese withdrew their main defensive line into a perimeter extending from western New Guinea through the Carolines to the vital Marianas. This line was to be held at all costs.¹⁹

Though the Marianas were very important to the Japanese, and Admiral King also considered them so, many planners in the Army and Navy had not been convinced that the Marianas needed to be taken. This was obvious in Quebec at the Quadrant Conference in August 1943 where the islands were not mentioned in the written plans of the Joint Chiefs.²⁰ This state of affairs would change as some major strategic alterations were brought about by the talks.

At Quebec the Combined Chiefs suggested timing the defeat of Japan within a twelve-month period after the fall of Germany. This concept would mean accelerating operations in the Pacific, and perhaps extending their reach.²¹ In line with this reasoning, the decision was made to bypass Rabaul. Also, with King again vocal on the subject, the Marianas were approved, somewhat halfheartedly, as a target. The Marianas were conspicuously absent from a proposed timetable for operations. King, however, was able to offer a modification of the timing of operations by inserting into Specific Operations in Pacific and Far East this statement: It may be found desirable or necessary to seize Guam and the Japanese Marianas, possibly the Bonins, in conjunction with the seizure of the western Carolines, and in particular with the attack on the Palaus. The Mariana—Bonin attack would have profound effects on the Japanese because of its serious threat to the homeland.²² (It should be noted here that though the Combined Chiefs considered Pacific operations, strategic control of this theater had in fact been turned over to the American Joint Chiefs.)²³

King gained an important ally at Quebec: General Henry H. Arnold, commander of the Army Air Forces. Arnold’s new weapon, the B-29, had been mentioned at Quebec, but its importance was not yet understood. This would shortly change. The Quebec decisions and the apparent alliance between King and Arnold indicated to MacArthur that he and the Southwest Pacific might soon be playing second fiddle to Admiral Nimitz in the Central Pacific. MacArthur intensified his opposition to the Central Pacific route.

On 4 October 1943 Brigadier General Laurence S. Kuter USA, of the Joint Staff Planners, wrote a memorandum concerning the B-29, pressing for the capture of the Marianas. In it he stated, Current Planning in the Pacific treats the seizure of the Marianas as a subordinate operation. . .,²⁴ He continued that if this was the case, the planners ought to forget about the Marianas, for the operation would serve no purpose. But the new B-29s had the capability of reaching and bombing Japan—from the Marianas.

Another paper, originated by the Joint War Plans Committee and titled Outline Plan For The Seizure of The Marianas, Including Guam, was also circulating in JCS upper echelons about this time. The paper predicted, as had Admiral King earlier (and both correctly), that the Japanese Navy would probably come out and fight in defense of the islands.²⁵

The next meeting of the Combined Chiefs (Sextant) was held in Cairo in late November-early December 1943. General MacArthur sent his chief of staff, Major General Richard K. Sutherland USA, to oppose any more Central Pacific operations. MacArthur, through Sutherland, argued that amphibious operations in that area would take too long to mount; land-based air support would be unavailable because of the distances involved; and carrier-based air would not be strong enough to maintain pressure on the enemy. Therefore, Nimitz’s forces should be used to support MacArthur.²⁶

On the other hand, proponents of an offensive through the Central Pacific argued that such a route would require fewer and longer moves; could bypass enemy strongholds far easier than an army tethered to the range of land-based fighters; would be more direct; would cut Japanese sea communications more effectively; would provide a base (the Marianas) for the B-29s; and would probably draw the Japanese Navy out for a decisive naval engagement.²⁷

Along with King, General Arnold brought his influence to bear on the Combined Chiefs on the matter of the Marianas. Basing the B-29s in China, as was then planned, promised to be a headache, and Arnold had little confidence in the Chinese’s ability to defend the airfields. The Marianas would be easier to defend and supply. And he assured the CCS that the B-29s could bomb Japan from Guam, Tinian, or Saipan.²⁸

The Sextant Conference brought about some far-reaching agreements pertaining to operations in the Pacific. Two documents were particularly important in this regard: Specific Operations for the Defeat of Japan and Over-all Plan for the Defeat of Japan. Specific Operations set up a timetable for execution. It was strictly for planning purposes and took into account the fact that circumstances might change drastically. It called for the invasion of the Marianas on 1 October 1944, with the start of B-29 attacks from there in December.²⁹ The Over-all Plan, which was approved in principle at Cairo, was used as a jumping-off place for later studies. This plan established the strategic concept within the Pacific.³⁰

The revised Combined Chiefs plan was sent to Nimitz and MacArthur on 23 December 1943. In it a two-pronged plan was formulated. Each prong would be mutually supporting, and each would be executed in conjunction with the other. One prong would be led by MacArthur up the back of New Guinea, through the Netherlands East Indies, and into the Philippines. Meanwhile, Nimitz would lead the Central Pacific prong, which would stab through the Japanese Mandated Islands. Both routes would converge on the Luzon–Formosa–China area in the spring of 1945. The Central Pacific route would be favored if there was any conflict in the operations.

The planners believed that the Central Pacific route offered the advantages of a more rapid advance toward Japan and her vital lines of communications, with the earlier acquisition of strategic air bases (i.e., the Marianas) closer to the Japanese homeland. Finally, the planners repeated the point that an advance in this area would be more likely to precipitate a decisive engagement with the Japanese Fleet.³¹

Based on the directive of 23 December 1943 Admiral Nimitz issued his tentative campaign plan, code-named Granite, four days later.³² This was quickly followed by another Granite plan on 13 January 1944. There were a few differences in the two plans. The sequence of operations was to be as follows:

In this formulation, an invasion of the Marianas would take place rather late in the year, and the U.S. forces would have to backtrack from the Palaus to attack the Marianas. This sequence was soon dropped, however. Nimitz also pointed out in Granite that a major fleet action, even if it delayed an operation for a time, would be very likely to speed up those that followed.³⁴

On 27 and 28 January 1944 senior officers of the South, Southwest, and Central Pacific commands met at Pearl Harbor to discuss and plan forthcoming operations. Before the meeting Admiral Towers had sent Admiral Nimitz a memorandum endorsed by Rear Admirals Charles H. McMorris and Forrest P. Sherman (Nimitz’s chief of staff and planning officer), opposing the Marianas operations. Towers pointed out that the Japanese could easily send bombers escorted by fighters through Iwo Jima and Chichi Jima to attack the southern Marianas, whereas American efforts towards Japan would have to be without fighter escort. The lack of harbors in the Marianas would be a point against these targets also. Towers recommended that the Pacific Fleet go by way of the Admiralties and the Palaus, bypassing Truk and the Marianas, and striking for the Philippines.³⁵ Thus, he agreed precisely with what MacArthur had been recommending.³⁶ Nimitz concurred with the recommendation and sent it on to King. There it would get a chilly reception.

During the meetings on the 27th and 28th Admiral Nimitz proposed two alternative timetables for future operations. These were:

(A) Truk, 15 June; Marianas, 1 September; Palaus, 15 November.

(B) Truk, bypass; Marianas, 15 June; Palaus, 10 October.³⁷

Most of the officers present were not enthusiastic about a Central Pacific offensive, and opposition to the capture of the Marianas was almost unanimous.³⁸

Towers presented his views again, and Sherman continued complaining about the poor harbors and the cost in lives of such an operation. Major General George C. Kenney, USA, who hoped to get B-29s under his control in the Southwest Pacific, thought that operating the big planes out of the Marianas was just a stunt.³⁹ With everyone present against a Marianas attack, Nimitz said he would recommend to Washington a single prong attack, using the New Guinea—Mindanao approach as the line of advance.⁴⁰

When he heard the results of the conference, MacArthur was elated, believing he had finally won the ballgame. But there was one heavy hitter yet to come to bat—Admiral King. MacArthur went about making his plans, including an obvious one to make himself the supreme commander of all forces in the Pacific. But with King yet to bat, MacArthur’s plans were presumptuous and his hopes were soon dashed.⁴¹

King meanwhile had written a blistering letter to Nimitz, saying in part:

Apparently, neither those who advocated the concentration of effort in the Southwest Pacific, nor those who admitted the possibility of such a procedure, gave thought nor undertook to state when and if the Japanese occupation and use of the Marianas and Carolines was to be terminated. I assume that even the Southwest Pacific advocates will admit that sometime or other this thorn in the side of our communications to the western Pacific must be removed. In other words, at some time or other we must take out time and forces to carry out this job. . . .

The idea of rolling up the Japanese along the New Guinea coast, throughout Halmahera and Mindanao, and up through the Philippines to Luzon, as our major strategic concept, to the exclusion of clearing our Central Pacific line of communications to the Philippines, is to me absurd. Further, it is not in accordance with the decisions of the Joint Chiefs of Staff . . . .⁴² (Italics added.)

King’s last sentence is crucial. No matter how much the commanders in the Pacific talked, without solid evidence against a Marianas invasion, the views of the JCS—particularly King and Arnold—would prevail.

Nimitz was already having second thoughts about a drive across the Central Pacific, having been buoyed by the successful Marshalls landings. Still, he was concentrating on Truk, not the Marianas. King was again not pleased, having himself eliminated Truk as a target some time before. He again prodded Nimitz toward a Marianas decision, saying, We should not do so [mount a Truk invasion] unless and until all other measures have been fully examined.⁴³

General MacArthur had kept up his fight against the Central Pacific route, dispatching General Sutherland to Washington in early February, in a last-ditch effort to change the Joint Chiefs’ minds. It was a futile effort, however, and MacArthur had to accept the inevitable. To the end, however, he believed that the Central Pacific route was the wrong way to defeat Japan.

Another visitor to Washington in March was Admiral Nimitz, also there to present his views and plans on upcoming operations to the JCS. The Marianas were discussed in detail, particularly regarding their lack of harbors. Eniwetok would be too small for full development and would be too far east for operations west of the Marianas. Nimitz mentioned Ulithi, about four hundred miles southwest of the Marianas, as a possible fleet base. (Ulithi was finally taken in September during the Palaus invasion and proved to be an excellent anchorage for the Pacific Fleet in its later operations.)⁴⁴

A speedup of operations was considered by the JCS, but Truk at first remained a stumbling block. Nimitz felt that an assault on Truk should be avoided if possible,⁴⁵ but this was a measure not to be taken lightly. Until the carrier strikes on Truk in February showed what to expect there, the original timetable had to be followed. When the TF 58 raids on 17 and 18 February showed that Truk was not the bastion it was thought to be, it was obvious that Pacific operations could be expedited.

Besides the Truk strikes, a number of factors entered into the Joint Chiefs’ decision to accelerate the offensive. The Marshalls operation had taken less time and had been less difficult than anticipated. MacArthur, perhaps hoping to get the jump on Nimitz and also show the JCS an attack in his area would be cheaper and faster, advanced the date of his invasion of the Admiralties from 1 April to 29 February and proposed bypassing Wewak and Hansa Bay to strike directly at Hollandia. With Truk obviously weak, Nimitz was now recommending an attack on the Marianas in mid-June, to be followed by an assault on the Palaus around 1 October. Finally, General Arnold was still pushing for the early capture of the Marianas to deploy the B-29S there sooner.⁴⁶

Taking into account these new factors, the Joint Chiefs issued a directive on 12 March 1944. In it MacArthur was ordered to cancel the Kavieng landings and instead isolate that island and Rabaul. Hollandia was to be targeted for 15 April 1944. All Pacific Fleet combat and service vessels, which had been allocated for the Kavieng–Hansa Bay–Manus attacks, were to be returned to Nimitz by 5 May. Following Hollandia MacArthur was to undertake further operations up the New Guinea coast and to prepare to assault Mindanao and support the Central Pacific attack on the Palaus.

In the same directive Admiral Nimitz was ordered to institute and intensify to greatest practicable degree bombings of the Carolines and to make carrier attacks on the Marianas, Palaus, Carolines, and other profitable targets.⁴⁷ The southern Marianas were to be seized, with a target date of 15 June 1944, and B-29 bases and secondary naval bases established. Nimitz was also ordered to cover MacArthur’s Hollandia attack and other operations in that area. Both commanders were ordered to prepare plans for the coordinated and mutual support of these operations.⁴⁸

The day after receiving the JCS directive Admiral Nimitz sent a secret dispatch to his major subordinates. In it he ordered that planning for the Truk assault be halted and all effort be given to preparing for the Marianas. A week later a CinCPOA Forager Joint Staff Study was issued. In it was outlined the purpose of the operation. Specifically, it was to establish bases from which to attack the enemy’s sea-air communications, support operations for the neutralization of bypassed Truk, initiate B-29 bombing of the Japanese home islands, and support further offensives against the Palaus, Philippines, Formosa and China. Generally, it was to maintain unremitting military pressure against Japan and to extend our control over the Western Pacific.⁴⁹

A new campaign plan, Granite II, was issued on 3 June. There was a marked difference between this and the original Granite plan, particularly in the timing of operations. The tentative schedule was now:

When this plan was issued the invasion of the Marianas was less than two weeks away. The forces on the long winding road to Tokyo were approaching